Tuesday, March 26, 2013

Selecting a therapy in CLL

CLL is unique because there is often a significant lag in time between when the diagnosis is made and the first treatment is needed.  Since complete blood counts (CBC's) are fairly routine, and CLL sticks out like a sore thumb, patients can often watch their disease for a long time before needing to start on treatment.  This can create a dangerous "paralysis by analysis" situation where careful consideration of all options leaves a patient unable to decide on what to do - the downsides of ANY treatment can look really big when they are staring you in the face.

When it comes to treatment decisions, I need to make one an argument for "watch and wait."  We are so conditioned to believe that early treatment improves outcomes in cancer that watch and wait feels very counterintuitive.  I have been really struck by several recent publications that evaluate "clonal evolution" in CLL.  I've written about them in two prior posts linked here and here

I could quickly sumarize by saying that CLL may start as only one clone, but it can often give rise to several clones over time.  If some of the clones that evolve over time are "smarter" (ie. have chemotherapy resistance, faster growth, predisposition to transform - see my post on the "new markers" and "CLL Prognosis") - it may be to your advantage to leave the "stupid" clone alone.

Treatment often eliminates the "stupid" clone and gives more room for the "smarter" clone to take over.  In CLL, we take it for granted that high risk markers like 17p, 11q, BIRC3, P53 mutations are more common in relapsed disease.  What we haven't considered until recently is that our treatment may be selecting for these higher risk markers.

So I am a big fan of "watch and wait" until treatment is needed.  Please see my post on "when to treat CLL" for more of a discussion about when watch and wait is no longer indicated.

But sometimes - we can't wait any longer.....

I've had a fun time serving on a leadership team for a large registry study called the "Connect CLL Disease Registry" sponsored by Celgene.  Other leaders on the study include Neil Kay, Michael Keating, Ian Flinn, Mark Weiss, Nicole Lamana, Chris Flowers etc.  We've collected a lot of data about how patients with CLL are treated throughout the United States.  We've reported some of our data at American Society of Hematology but we've really only just gotten started. 

In the United States, most patients treated off of a clinical trial are going to receive either FCR (Fludarabine, Cyclophosphamide, Rituxan), FR (get rid of the cyclophosphamide), or BR (bendamustine, rituxan).  Chlorambucil is a respectable option for the elderly population with other medical issues but is generally reserved for patients you just can't treat with chemotherapy (editorial - I think chlorambucil is underutilized as a treatment option in elderly patients - just telling you what the patterns are out there).  There is also a surprising amount of single agent rituximab use out there. 

Thus far, there have been NO PUBLISHED REPORTS that compare FCR to BR in a prospective randomized study (Gold standard for evidence).  The study is being done by the Germans but we don't expect data from that study soon.  Chaya Venkat did a very thoughtful evaluation on her blog CLL Topics looking at the data from the single arm phase II study of BR in the front line.  I would direct the motivated reader to her post. 

To be very clear - when evidence is lacking - opinion abounds.  Right now in 2013 we do not know what the best front line therapy actually is.  FCR has been the leader of the pack for a number of years, but there are a lot of reasons FCR isn't always the right answer for all patients. 

Without a clear answer - opinion is fair game - I will share mine.  My first choice for therapy is to consider participation in a clinical trial.  That is the "standard answer" in the NCCN guidelines, but it is more true in CLL today than just about any other disease.  New investigational treatments like ibrutinib, idelalisib, GA-101, ABT-199, TRU-016 are exploiting the biology of CLL and bringing forward a new generation of treatments for patients with CLL that are currently only available in clinical trials.

The Germans have wonderful terminology that segregates patients into different groups of patients called the "go-go" population, the "slow go" population and the "no go" population for the patients who are young and fit vs the patients in whom you need to have more caution, vs the patients that have a ton of medical issues and the goal is to do no harm.  They use a tool called the "cumulative illness rating scale" aka CIRS which has actually been around for quite a few years but gives you a tool to mathematically quantify how many medical problems other than CLL a patient has.  The sicker you are - the less intensively you can treat the CLL.

I look at a patient and ask myself if they are "fit" for a fludarabine based regimen.  Fludarabine is a good drug - but it has some potential draw backs.  It is cleared by the kidneys, so if the kidneys are less than ideal - problems start to increase.  Since kidney function declines with age, and CLL is more common in the elderly, the kidneys often guide my decision making.  Kidney function can be measured by the "creatinine" which is a standard measure on a chemistry panel.  Take age, gender, creatinine and you can calculate a "glomerular filtration rate" or GFR.  When that number dips below 60, my caution level goes up quite a bit.

So what problems can fludarabine cause?   It can lower the immune system, cause prolonged supression of the bone marrow, and worsen a hemolytic anemia.  Therefore, in a patient with bad lungs from smoking (COPD) who has had pneumonias, has a GFR well under 60, or maybe already has signs of the immune system attacking the red blood cells or the platelets - I avoid fludarabine based chemo and lean toward bendamustine based treatment.

If a patient is "fit" for fludarabine, the choice is either FR or FCR.  There is no question that the addition of the cyclophosphamide adds a punch - both to the CLL and to the patient.  The Germans have published a study looking at F vs FC (pre-rituximab era).  This was also studied in the US by the Eastern Cooperative Oncology GroupThe English also looked at this in the CLL4 study.    The English did a good job looking at biomarkers in their effort and came to the conclusion that patients with the 11q deletion did a better when they had the cyclophosphamide added to fludarabine.  There has been some debate whether the cyclophosphamide is necessary in the absence of 11q deletion.  I tend to reserve it for the particulary fit younger patient or the patient with the 11q but I acknowledge this is an opinion and there are plenty of bright CLL docs that would disagree and give it more broadly.  It goes without saying that any regimen would include rituxan as shown to us by the German CLL study group. 

If a patient is "unfit" for a fludarabine regimen the decision in my mind comes down to Bendamustine versus chlorambucil.  Bendamustine is more effective than chlorambucil but perhaps a little more intensive.  Bendamustine does not require the kidneys for elimination - so when GFR is a little lower, this can be a good option.  For an elderly patient, chlorambucil may still be a good option and should not be disregarded completely.  Our patterns of care study indicated that it is the treatment decision of only about 10% of patients over the age 75 and much less common in younger individuals.  If I give bendamustine, I virtually always give it with rituximab.

There are some special populations to consider.  The biggest one in my mind is the 17p deleted group.  We know that cells with the 17p deletion don't respond well to DNA damaging chemotherapy (FCR, BR, etc).  I always want to know, "how many 17p deleted cells they have?"  Keep in mind that FISH testing gives you a rough percentage of cells with a particular marker.  A patient with 5% 17p deleted cells is very different to me than a patient with 85%.  The first patient is likely to respond reasonably well to treatment but relapse earlier with a disease that is predominantly 17p deleted.  The latter patient is likely to respond very poorly to FCR (short duration of response or limited response at all).Pretty soon we will also be looking at SF3B1 and BIRC3 in this setting and asking the same questions.

For a patient with a large population of 17p deleted cells, there are regimens out there now that use high doses of steroids in combination with either rituximab or campath.  We know that campath is one of the few drugs that largely ignores the 17p deletion status but it is not a drug that is easy to use.  It definitely lowers the immune system quite a bit and you need to be careful about infections with things that you don't always consider (PCP, CMV, shingles, etc.).  Most docs who use campath give preventative antibiotics for each of these - which can definitely be a mouthfull of pills.  Campath has recently been taken off the market - but you can still get it.  It is a little cumbersome (ie. given three times per week for 12-16 weeks) but it can be quite effective in this setting.  Frankly, if I can buy a patient a year or two without really messing up their genome, I think I will be able to get them some of the research drugs that appear quite effective against 17p deletetion CLL including ibrutinib or ABT-199.

Of course, not many patients are appropriate for allogeniec stem cell transplant, but these might be the folks.  I am also really intrigued by the "engineered T-Cell" therapies being developed that may be a good option for these patients that don't respond well to typical chemotherapy.

That is the basics of my thinking for patients with untreated CLL that need some form of treatment.  This is no substitute for the advice of your own doc, but at least how I think through the issue.  There are a number of modifications to what I wrote that would be entirely acceptable.  With any luck this will all be outdated in a few years as new options come on line. 

 Hope that helps.